Details
- Vendor
- Twitter (now X)
- Type
- Other
- Category
- Established
- Status
- Inactive — Archived and unmaintained since 2017. Twitter is no longer responding to issues or PRs.
- License
- Apache-2.0
- Language
- Scala
- Query Languages
- Thrift API
- Protocols
- Thrift
FlockDB
A distributed graph database built by Twitter for storing adjacency lists, optimized for wide and shallow social graphs like follower relationships. Supports set arithmetic queries but intentionally avoids multi-hop graph traversal.
- Overall rank
- #67
- First released
- April 2010
- License
- Apache-2.0
- Built with
- Scala
Details
Sources 4 references
Protocols Thrift
Twitter's 'flapp' servers expose a small Thrift API (default port 7915 in the demo); the official Ruby client wraps it, but any Thrift client can talk to it directly.
-
The flockdb interface is thrift, so you can talk to it in many different languages
Flock demo · reviewed 2026-08-02
License Apache-2.0
Confirms current value; the repo's LICENSE file is the Apache License 2.0 text, matching GitHub's 'Other (NOASSERTION)' classification quirk (Apache headers without the SPDX-standard NOTICE format).
-
Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License.
LICENSE · reviewed 2026-08-02
Released 2010-04
Corrects current year-only '2010' to 2010-04: the repo was created 2010-04-12 and its own README already describes FlockDB as being in full production use that same month.
-
As of April 2010, the Twitter FlockDB cluster stores 13+ billion edges and sustains peak traffic of 20k writes/second and 100k reads/second.
twitter-archive/flockdb: A distributed, fault-tolerant graph database · reviewed 2026-08-02
Language Scala
Confirms current value.
-
The app servers (affectionately called "flapps") are written in Scala, are stateless, and are horizontally scalable.
Introducing FlockDB (doc/blog.md) · reviewed 2026-08-02